Collect blind feedback from a mock jury
For mock trial facilitators, instructors and advocates evaluating a presentation.
How do you find out what listeners actually understood without giving away the answer you hope to hear?
Keep the presentation goal and the advocates’ preferred conclusion from the participants. Collect individual comments before group discussion, ask for an explanation in their own words and preserve the distinction between first impressions and later reactions.
Decide exactly what you are testing
Choose one question a presentation can answer, such as whether listeners understand the sequence of two inspections. Use a fictional or otherwise authorized packet. If comparing versions, change one feature, such as the order of two exhibits, and keep the presenter’s time and instructions comparable. Assign participants to versions by chance where practical.
Use neutral version labels and omit clues about the preferred result. Tell participants what facts and instructions to use, whether they may take notes and when discussion begins. Record anything they already know about the exercise or its organizers that limits blinding.
Collect independent explanations first
Before offering a multiple-choice answer or opening discussion, ask each participant to write what they understood, which item mattered and what remained unclear. Then ask a specific factual question that can be scored consistently. Give “cannot tell from the packet” as an option when appropriate. Collect responses without names if the design permits it.
Save the individual answers before the group talks. After discussion, ask what changed and which evidence or explanation prompted the change. Record the two stages separately. Pew Research Center’s survey-methods guidance explains how question wording and order can affect responses. This exercise applies that principle to mock feedback.
Use disagreement to improve the next round
Count the responses by version and describe how many listeners participated in each group. Read the actual reasons. One person’s unexpected but textually supported interpretation may reveal an ambiguity a vote total would hide. Small convenience groups can help identify a confusing exhibit or presentation choice. They cannot reliably predict what a real jury will decide.
Worked example
Illustrative debrief. Of six listeners shown timeline A, four say the second inspection happened before the warning. Of six shown timeline B, one gives that answer. Ask both groups which exhibit led to the belief before concluding the sequence is clearer in B. Five listeners point to the same undated image. The team’s next test labels that image and repeats the exercise with new participants.
Primary sources
Related work at Agentic Attorney
This mock-jury exercise is part of the shared Agentic Attorney resource collection. The main site describes its consulting and implementation work for legal teams. Explore Agentic Attorney’s approach to AI implementation.